flagged as dupe post when string is unique (ideas?)

[gembry]

Hello Folks, (fishing for input) thx

Not really sure what’s going on here. Below you will find what I am
sending and what is coming back from the Twitter API. You can see that
what is being sent doesn’t even come close to what is up there now.

Am I missing something here? What am I doing wrong? Why is this
completely different sting not posting?

>>> Sending >>>

method="post"
url="http://twitter.com/statuses/update.xml?status=Great%20Job%20for
%20OAS%20TEST%20ONLY%20in%20California%2DAnaheim%2FHuntington%20Beach
%2E%20http%3A%2F%2Flocalhost%2FAssignmentReady%2Findex%2Ecfm%3ForderID
%3D2315510%26category%3DAllied%5FTravel%26ts%3D102809140819%20or
%20800%2E469%2E5314"

<<< Result (from Twitter API) <<<

<?xml version="1.0" encoding="UTF-8"?> <status> <created_at>Wed Oct 28
19:51:30 +0000 2009</created_at> <id>5238791820</id> <text>what</text>
<source>web</source> <truncated>false</truncated>
<in_reply_to_status_id></in_reply_to_status_id>
<in_reply_to_user_id></
in_reply_to_user_id> <favorited>false</favorited>
<in_reply_to_screen_name></in_reply_to_screen_name> <user>
<id>57127952</id> <name>Larp Master Flash</name>
<screen_name>LarpMasterFlash</screen_name> <location></location>
<description></description> <profile_image_url>http://s.twimg.com/a/
1256674706/images/default_profile_0_normal.png</profile_image_url>
<url></url> <protected>false</protected> <followers_count>4</
followers_count> <profile_background_color>9ae4e8</
profile_background_color> <profile_text_color>000000</
profile_text_color> <profile_link_color>0000ff</profile_link_color>
<profile_sidebar_fill_color>e0ff92</profile_sidebar_fill_color>
<profile_sidebar_border_color>87bc44</profile_sidebar_border_color>
<friends_count>0</friends_count> <created_at>Wed Jul 15 20:31:16 +0000
2009</created_at> <favourites_count>0</favourites_count>
<utc_offset></
utc_offset> <time_zone></time_zone>
<profile_background_image_url>http://s.twimg.com/a/1256674706/images/
themes/theme1/bg.png</profile_background_image_url>
<profile_background_tile>false</profile_background_tile>
<statuses_count>34</statuses_count> <notifications>false</
notifications> <geo_enabled>false</geo_enabled> <verified>false</
verified> <following>false</following> </user> <geo/> </status>

Thanks for the input.

[John Kalucki]

Is the posted status similar to any other status created by that user?

-John Kalucki
http://twitter.com/jkalucki
Services, Twitter Inc.

[Andy Freeman]

> Is the posted status similar to any other status created by that user?

Does the above imply that similar will trigger the dup detector?

Argh!

Please don't tell me that you're now rejecting similar tweets....

Url shorteners can easily generate similar urls, so if someone is in
the habit of tweeting "check this out" followed by a shortened url,
the tweet text is likely to be similar even though the actual url is
different.

Heck, urls that are similar often point to very different web pages.

There are lots of other cases where similar tweets can actually be
very different, and thus should not be rejected as "duplicates".
(I've heard of a project that tweets data that is guaranteed to be
unique but is reasonably likely to be similar because of aggressive
encoding.)

On Nov 1, 6:22 am, John Kalucki <jkalu...@gmail.com> wrote:
> Is the posted status similar to any other status created by that user?
>

> -John Kaluckihttp://twitter.com/jkalucki

> > Thanks for the input.- Hide quoted text -
>
> - Show quoted text -

[Raffi Krikorian]

>> Is the posted status similar to any other status created by that
>> user?
>
> Does the above imply that similar will trigger the dup detector?
>
> Argh!
>
> Please don't tell me that you're now rejecting similar tweets....
>
> Url shorteners can easily generate similar urls, so if someone is in
> the habit of tweeting "check this out" followed by a shortened url,
> the tweet text is likely to be similar even though the actual url is
> different.
>
> Heck, urls that are similar often point to very different web pages.
>
> There are lots of other cases where similar tweets can actually be
> very different, and thus should not be rejected as "duplicates".
> (I've heard of a project that tweets data that is guaranteed to be
> unique but is reasonably likely to be similar because of aggressive
> encoding.)

as noted by other people on this list, twitter is currently rejecting
tweets that match either your last update, or an update you recently
sent. unfortunately, the API is currently silently failing but it is
on the short list to have the API return an error code instead.

--
Raffi Krikorian
Twitter Platform Team
ra...@twitter.com | @raffi

[Andy Freeman]

> as noted by other people on this list, twitter is currently rejecting
> tweets that match either your last update, or an update you recently
> sent. unfortunately, the API is currently silently failing but it is
> on the short list to have the API return an error code instead.

Can you confirm what "match" means? Is it "same character sequence"
or is it "similar character sequence"?

I ask because John Kalucki wrote "Is the posted status similar to any
other status created by that user?" above in response to a question
about duplicate rejection.

Thanks,
-andy

> ra...@twitter.com | @raffi- Hide quoted text -